If At First You Don’t Succeed…

January 14, 2010 at 2:04 pm

A consultant at Mayo Clinic Rochester requested a search for patient care.  The request he made was:

“Any examples you can find where a change (increase) from a baseline level of a test is used to confirm a diagnosis.”

Thinking this was a little broad, I asked him for some clarification.  He replied:

“I was looking for something like a mediator, hormone, cytokine that spikes when symptoms flare.”

This gave me something manageable to work with.  Plus, having done many searches for this physician in the past, I knew he was most likely interested in allergic diseases, as that is his specialty.

In the OVID Medline database there are the subject headings: Inflammation Mediators, Hormones, and Cytokines.  I also knew that the subheading “Diagnostic Use” was available for each of these subject headings.  The scope note for Diagnostic Use defines it as:

Subheading Information for Diagnostic Use
Used with chemical compounds, drugs, and physical agents when used for studies of clinical function of an organ, or for the diagnosis of human or animal diseases.  Year of entry:  1967.

So I combined all these items together in one set with the following search:

Set #  Search

1     exp Inflammation Mediators/du [Diagnostic Use]
2     exp Hormones/du [Diagnostic Use]
3     exp Cytokines/du [Diagnostic Use]
4     1 or 2 or 3

Where “exp” is the notation denoting that the subject heading has been “exploded” to capture all the narrower subject headings in the set.  For example, the subject heading Hormones has the narrower subject headings Androgens, and Estrogens, among others.  Exploding Hormones retrieves all the citations for these narrower subject headings in addition to the citations indexed under the broader subject heading Hormones.

Then I narrowed the results down to those references dealing with allergic diseases.  In Medline the subject heading for this topic is Hypersensitivity.

Set #   Search

5     exp Hypersensitivity/
6     4 and 5

Much to my dismay, however, the resulting set was quite unsatisfactory.  There really was very little that seemed relevant to the physician’s topic.  This was confirmed by him from the few “best results” I sent for him to look over.  I mentioned this to a colleague and she suggested I try the subheading “Blood”, as the diagnostic tests would be testing blood levels, as well as combining with the subject heading “Reference Values” which would capture the concept of baseline in the requester’s original statement.  Maybe using Reference Values would find citations comparing baseline values to abnormal levels.

The definitions for these two are:

Subheading Information for Blood
Used for the presence or analysis of substances in the blood; also for examination of, or changes in, the blood in disease states.  It excludes serodiagnosis, for which the subheading “diagnosis” is used, and serology, for which “immunology” is used. For the presence of microbes or parasites in the blood in disease, use “microbiology” or “parasitology”.  Year of entry: 1967.

MeSH HEADING: REFERENCE VALUES
SCOPE: The range or frequency distribution of a measurement in a population (of organisms, organs or things) that has not been selected for the presence of disease or abnormality.
NOTE: never IM; no qualif; differentiate from REFERENCE STANDARDS: REFERENCE VALUES = range or frequency distribution of a measurement, REFERENCE STANDARDS = basis of value or standard for a measurement YEAR of ENTRY: 80; was NORMAL VALUES 1977-1979

Using these concepts produced the following search:

Set #   Search

7      exp Inflammation Mediators/bl [Blood]
8      exp Hormones/ bl [Blood]
9      exp Cytokines/ bl [Blood]
10    7 or 8 or 9
11    Reference Values/
12    5 and 10 and 11

Unfortunately this search also yielded very little.  The resulting set was small and nothing seemed on topic.  I immediately suspected Reference Values was limiting the search too much.  I discarded that topic, performing a search on Hypersensitivity and blood values (5 and 10).  Further limiting to English Language, Humans only, and articles published in the last five years produced a nice set of a few hundred hits.  Further browsing of this set allowed me to discard the obviously irrelevant citations to provide a result the physician was happy with.

In conclusion, it is easy to see that searching isn’t an exact science.  Sometimes the obvious path to search results leads you astray, and it takes several refinements to your search strategy before the desired results reveal themselves.  Often talking over your strategy with someone either knowledgeable with searching and/or the subject area provides clues for further sleuthing.

Larry Prokop, Reference Librarian, Mayo Clinic Libraries

Entry filed under: What's Your Reference Question?. Tags: .

Professional Travel & Meeting Attendance NLM Updates



%d bloggers like this: